AI written...
AI Slop; Did Not Read
We rebuilt the Linux MicroVM stack on Apple Silicon
71–80 of 88 posts
Re: We rebuilt the Linux MicroVM stack on Apple Silicon
#72Earlier quoted context omitted.
> Buy any laptop, stick Linux on it, get working. Except it's still a slightly different challenge each time to "get it working", each different PC. A Mac is a Mac is a Mac, more or less. I have a nice Framework laptop on my desk running Kubuntu alongside my Macs. Every now and then I check in, make sure I could switch to it if I needed to, test all my important tasks, plug it in so the stupid mainboard battery does…
> Except it's still a slightly different challenge each time to "get it working", each different PC Maybe thirty years ago. Not today. Just go and pick up a second-hand Thinkpad for pennies. Everything will work on it.
Right but a secondhand thinkpad is not a laptop you find in the travel context of the thread.
(I dispute the main point, too — I spent flipping hours trying and failing to get trackpad speed right on a Surface Go in GNOME/Wayland. Ended up having to abandon that combination.)
Re: We rebuilt the Linux MicroVM stack on Apple Silicon
#73Earlier quoted context omitted.
> Except it's still a slightly different challenge each time to "get it working", each different PC Maybe thirty years ago. Not today. Just go and pick up a second-hand Thinkpad for pennies. Everything will work on it.
Right but a secondhand thinkpad is not a laptop you find in the travel context of the thread. (I dispute the main point, too — I spent flipping hours trying and failing to get trackpad speed right on a Surface Go in GNOME/Wayland. Ended up having to abandon that combination.)
> Surface Go
Avoid.
Lenovo only, T or P series. At least in the samples I've had, all the drivers have been in-kernel upon release of the hardware.
Re: We rebuilt the Linux MicroVM stack on Apple Silicon
#74Earlier quoted context omitted.
Right but a secondhand thinkpad is not a laptop you find in the travel context of the thread. (I dispute the main point, too — I spent flipping hours trying and failing to get trackpad speed right on a Surface Go in GNOME/Wayland. Ended up having to abandon that combination.)
> Surface Go Avoid. Lenovo only, T or P series. At least in the samples I've had, all the drivers have been in-kernel upon release of the hardware.
On the contrary it (faster Surface Go 2) was one of the most practical little machines I’ve ever used, and was really impressive in Windows 11.
I am totally happy to lay the blame for the poor trackpad performance in Gnome/Wayland on the software, because the hardware was super. Absent that problem I would have used it for almost anything and I loved taking it on trains and to coffee shops; Apple has no equivalent machine at all.
But you’re making the point for me. Success in Linux involves either tinkering or constrained hardware choices, even now. That’s fine but I think the Linux world pretends the situation is better.
Re: We rebuilt the Linux MicroVM stack on Apple Silicon
#75"You can even get /dev/kvm inside a Linux VM on an M3 or later running macOS 15, which runs Firecracker unmodified." I'm very confused why any SWE is on a M1 or M2 Mac. I've always gotten a new machine at $CORP every 2 years...
I have both an M1 Max and an M4 Max and out of habit am still on the former. The performance is still substantial enough that is quite usable.
Re: We rebuilt the Linux MicroVM stack on Apple Silicon
#76Earlier quoted context omitted.
> Surface Go Avoid. Lenovo only, T or P series. At least in the samples I've had, all the drivers have been in-kernel upon release of the hardware.
On the contrary it (faster Surface Go 2) was one of the most practical little machines I’ve ever used, and was really impressive in Windows 11. I am totally happy to lay the blame for the poor trackpad performance in Gnome/Wayland on the software, because the hardware was super. Absent that problem I would have used it for almost anything and I loved taking it on trains and to coffee shops; Apple has no equivalent ma…
"Success in MacOS” also has constrained hardware choices :)
> That’s fine but I think the Linux world pretends the situation is better.
The only thing comparable is Microsoft. Nothing else comes close to trying to support the variety of hardware and software from different places.
Re: We rebuilt the Linux MicroVM stack on Apple Silicon
#77I don't use MacOS.
I don't use MacOS because it is not Open Source.
I don't use Windows.
I used Windows extensively in the past, but I don't use it in 2026, because it is not Open Source.
Anytime you're dealing with a closed source OS, you're dealing with the following things:
1) Extra time/effort/complexity to make your programs work correctly.
2) Extra time/effort/complexity to make your already working working programs keep working through OS version updates, upgrades, patches, etc., etc.
This article (AI written, not AI written, I don't care) describes something that is overly complex (proof that you don't have and/or cannot access the source code of the underlying operating system) and something which could potentially fail when Apple releases its next OS or OS upgrade...
If you switch to an Open Source operating system (and I'm not trying to shill for Linux here -- OpenBSD, etc., any open source Operating System could work too), then most of these problems will disappear -- or at least be reduced rather significantly.
But what I'm most concerned about is the complexity...
Now, you're welcome to engineer whatever additional complexity above Apple's stack -- many companies do this and are more than willing to pay the price in additional time, resources, programmers, tech people handling whatever issues arise, as they arise.
I'm not against Apple -- I think their products have beautiful aesthetics, and every company, every producer of physical products on the planet could take a lesson from their design teams. Their products are absolutely stunningly visually appealing!
But Apple in turn, could take a lesson from Linux and every other open source OS on the planet.
I personally will not develop for any Apple product unless and until Apple open sources their OS'es.
If code that I write coincidentally runs on an Apple product, hey that's a happy coincidence! Yay!
But I never expect it to, and I never expect to keep working across OS versions and upgrades.
If anyone thinks I'm being unfair to Apple, then I'll state for the record that I have the exact same expectations (or negative expectations, as the case may be!) about Microsoft Windows!
Closed source OS'es lead to complexity, work-arounds ("oh no, API call isn't working the way we thought it would under these conditions -- well, use this work-around and call API with this unnecessary and overly complex data structure after making 3 other unnecessary API calls and then you need to do the following additional steps..."), and software breaking over time over OS version upgrades, updates, patches, and the like.
On the one hand, I'm very glad that this article is bringing micro VM's to Apple.
That's a good idea!
(But on the other hand, I have to ask if the complexity is worth it... but I'm not the guy who pays these guys' salaries or the future support guys' salaries, so I can't determine if it is or isn't, if the trade-off is worth it!)
On the positive side though, bringing Micro VM's to Apple is a good idea, at least in its purest idea form ("we want to add functionality X to platform Y" -- there's nothing wrong with the intention to add value to something -- in fact, quite the opposite! The whole of Western Civilization is based on that idea!)
Re: We rebuilt the Linux MicroVM stack on Apple Silicon
#78A suggestion to the author -- don't outsource your writing to AI, or, if you do, go over it with a fine-tooth comb and remove every trace of Claudisms from it. I read the subtitle: "What it takes to boot the same microVMs on Apple's hypervisor, and the one capability Apple will not let you have." and immediately navigated away.
Interestingly, the content has changed since I looked at it a few hours ago. For one, the cringy clearly AI generated intro sentence is refined.
Re: We rebuilt the Linux MicroVM stack on Apple Silicon
#79> most engineers at Encore develop on a Mac They are solving the wrong problem.
Hacker News is so utterly detached from reality at times, and comments like these really drive that home.
Hacker News is?
I worked on an internal tooling team at the beginning of the Apple Silicon transition; we spent months fixing ARM runtime issues that none of our customers encountered. That company probably spent millions of dollars fixing bugs that would never appear in production.
Re: We rebuilt the Linux MicroVM stack on Apple Silicon
#80Earlier quoted context omitted.
> Surface Go Avoid. Lenovo only, T or P series. At least in the samples I've had, all the drivers have been in-kernel upon release of the hardware.
On the contrary it (faster Surface Go 2) was one of the most practical little machines I’ve ever used, and was really impressive in Windows 11. I am totally happy to lay the blame for the poor trackpad performance in Gnome/Wayland on the software, because the hardware was super. Absent that problem I would have used it for almost anything and I loved taking it on trains and to coffee shops; Apple has no equivalent ma…
> But you’re making the point for me. Success in Linux involves either tinkering or constrained hardware choices, even now. That’s fine but I think the Linux world pretends the situation is better.
Have you heard of this new Windows thing that everyone's going on about? Almost nothing supports it out of the box and it's just all tinkering - and that's for something you have to pay a lot of money for.
You're equally hardware-constrained with Apple too. Can you find me an Apple laptop with four "proper" USB ports, a Displayport socket, and swappable battery?